In or der to solve the problems of high temperature corrosion and wear the project involves the development of an intermetallic iron-aluminium alloy (25-28% by mass of aluminium), whose metallurgical properties will have been improved by adding other elements and hot working processes. The ingot route and the powder route will be examined and atomization will be used to produce powders. Different coatings will be prepared too (PVD, semi-transferred and low pressure plasma spray). The behaviour of these alloys will be characterized in corrosion atmosphere (carburizing and sulphidizing environment). Wear tests will be conducted.
In order to solve the problems of high temperature corrosion and wear of steels the project involves the development of an intermetallic iron aluminium alloy (25 to 28% by mass of aluminium), whose metallurgical properties will have been improved by adding other elements and hot working processes. The ingot route and the powder route will be examined and automisation will be used to produce powders. Different coatings will be prepared too (physical vapour deposition (PVD), semitransferred plasma spray and low pressure plasma spray). The behaviour of these alloys will be characterised in a corrosion atmosphere (carburising and sulphidising environment) and wear tests will be conducted.
Iron aluminium (25% aluminium) with the addition of boron, zirconium and cerium have been prepared by both ingot route (vacuum casting) and by powder metallurgy. They have been extruded at 1100 C or formed by compression. Coatings of iron alluminium by plasma techniques and PVD have been prepared. Best adhesion and resistance to corrosion were obtained with PVD coatings.
In tensile tests, the alloys are brittle up to 400 C, above 600 C ductility increases. A good toughness (charpy V) is found for bimodal structure produced during compression.
A special reinforcement technique of the material results in a marked improvement of the creep resistance. Initial results indicate that this alloy exhibits higher resistance to sulphidising and oxidising corrosion, wear and creep, than stainless steel and is competitive with nickel based superalloys with the advantages of low density and low inherent material cost.
Probabilistic methods have been applied during the design stage in some industries to determine the life expectancy of a structure. Prior to this research, however, no method was available of taking into account the new information gathered during the operating phase of the structure to assist this probabilistic approach. The objective of this research was, therefore, to devise a method which would address the problem of including the effect of time and gross errors in the reliability study of a large multielement structure. From these considerations, the maintenance plan for the structure could be optimized by trying to minimize the inspection costs within reasonable constraints placed on the allowable probabilities of failure.
The research included the following elements:
the methodology was mathematically formulated to accommodate multielement systems;
the failure criteria was fatigue crack propagation;
data pertaining to the nature of the loads exerted, material properties, fatigue analyses on the typical components studied and on gross errors at the design and operating stage of the structure were organized into a common data base management system;
The time variant formulation of the reliability problem was then successfully implemented;
a survey of inspection costs and techniques, of maintenance procedures and an estimate of the consequences of gross errors was then undertaken to minimize inspection costs;
a parametric study in the overall maintenance optimization scheme was performed on a tubular intersection on a North Sea platform and illustrated in a satisfactory manner the application of the techniques and software developed;
a prototype expert system tool, capable of guiding a knowledgeable user in the decision making process and in managing different environments was the final outcome of the research;
a module, designed to analyze the effects of gross errors on the computation of maintenance costs was delivered as a stand alone expert system .
The objective of the research was to devise a database for predicting the maintence requirements of multielement structures which includes data from the operating phase of the structure. The first step consisted of setting up a database organizing all the information necessary for the reliability analysis of the structure. Data pertaining to the nature of the loads exerted, material properties, fatigue analyses on the typical components studied and on gross errors at the design and operating stage of the structure were therefore collected and organized into a common database management system. One of the objectives of the research was to implement software on personal computers (PC), therefore a new database was created. The Data Base manager DBASE III PLUS was chosen because of its availability, widespread use and affordable price. Furthermore it proved easy to interface with an expert system shell. The logical structures of the databases for material properties and mock up test data were then defined.
Since the subjects of inspection and maintenance are closely linked to the problem of fatigue, a special effort was directed towards the statistical assessment of fatigue and crack propagation data derived from test programs on laboratory specimens and mock-up geometries. Also relevant to the study was data pertaining to gross errors. The approach was to record all events including near misses and errors detected during inspection. The most relevant data available were found in the literature. With some additional contribution from industry it was possible to include 22 events from the nuclear industry and 20 events from the offshore industry in the database.
The operating environment is as follows :
Database manager: DBASE III PLUS
The research aimed to develop general structural reliability methods which are capable of handling time variant reliability factors where it is necessary to compute the probability that an adverse state is reached by the system for the first time. Based on studies carried out in parallel at the Technical University of Muenchen an innovative approach was applied in the course of the project to answer a need felt both in the nuclear and in the offshore industries. In these fields, the structures studied are constantly subjected to environmental loading, fluid transients and deteriorating structural properties, all of which are time variant phenomena.
A methodology to compute the probability of failure of a structural system at a reduced computer cost was developed and the impact of new observations on overall reliability prediction was determined by using Bayesian updating techniques. Though the formulation is general and may be used for a structural system, implementation was carried out for a single component. This limitation originated in the large programming effort required to model even a single component and from the absence of reliable data and information on the time dependent correlation of events concerning 2 neighbouring components which constitute part of the whole structural system.
The objective of this research was to produce an inspection and maintenance optimization package, based on the work performed on the formulation of the time variant reliability problem, augmented with information gathered on inspection strategies. A survey of inspection costs and techniques and of maintenance procedures was undertaken to provide input for the optimization scheme. The cost was minimized under a number of constraints including the statement that the reliability index must be larger than a minimum accepted value. The software developed delivers an inspection plan for a given component, in terms of when to inspect, how to inspect, when to repair and how to repair. A first estimate of the minimum overall inspection and maintenance cost of the structure is given at installation time. From this initial time on, complementary information about the real performance of the installation and the environmental conditions are gained both at inspection times and in between inspections. This new information is used in the software to update the uncertainties associated with the different parameters chosen to represent the sequence of events. A new inspection plan is obtained after every update. The resulting numerical package driving and encompassing the necessary numerical modules was called the INSPOP (inspection strategy optimization for offshore structures procedure) module. The numerical tools proved to work in a satisfying manner in an application to a real case.
The goal of the research was to define the general architecture of an expert system for the inspection and maintenance of industrial structures and to test the functionality of the interfacing between numerical applications and decision making processes. These goals were met in the creation of a prototype expert system encompassing all the features of a more complete tool. In order to reach this objective, knowledge mechanisms and main functions were also identified. The following results were obtained:
the feasibility of a large scale expert system working on a personal computer (PC), using a commercial shell such as Goldworks, was clearly demonstrated;
a prototype expert system (RAMINO) with the complete architecture, the main inference engines and part of the basic relevant knowledge was implemented and tested;
a complete selfconsistent module (GERM) for the analysis of gross errors with potential applications in many different domains was developed;
the interface problems, at the boundary of different programming environments (DBASE 3+, LISP, FORTRAN) were identified and a workable approach for their integration was designed and tested;
application software dealing with material modelling, stress modelling, general reliability analysis were organized and included in the system in a rational way, taking into account the interactions of the calculation modules with each other and with the database.
THE PRESENCE OF ALUMINIUM IN STEELS HELPS TO IMPROVE THEIR CORROSION RESISTANCE, BUT ALSO ADVERSELY AFFECTS THE METALLURGICAL PROPERTIES OF MATERIALS. IN ORDER TO SOLVE THE PROBLEMS OF HIGH-TEMPERATURE CORROSION AND WEAR THE PROJECT INVOLVES THE DEVELOPMENT OF AN IRON-ALUMINIUM ORDERED ALLOY (25-28% BY MASS), WHOSE METALLURGICAL PROPERTIES WILL HAVE BEEN IMPROVED. THE ALLOY WILL BE IN BOTH SOLID AND COATING FORM.
IN ORDER TO ACHIEVED THIS AIM, IT IS PLANNED, TO ADD CHROMIUM (0-5% BY MASS), BORON (LESS THAN 500 PPM) AND RARE EARTHS (LESS THAN 1%) TO THE ALLOY. THE ALLOY PREPARATION AND FORMING TECHNIQUES WILL BE DEVELOPED AND ATOMIZATION WILL BE USED TO PRODUCE POWDERS ENABLING EITHER SINTERED PRODUCTS OR COATINGS (PLASMA-ARC, SEMI-TRANSFERRED, LOW-PRESSURE PLASMA). THE BEHAVIOUR OF THESE HOT ALLOYS (SOLIDS OR COATINGS) WILL BE CHARACTERIZED IN BOTH AN OXIDIZING CORROSIVE ATMOSPHERE AND IN A LOW-OXIDIZATION, BUT CARBURIZING AND SULPHURIZING ATMOSPHERE. TESTS SHALL ALSO APPLY TO MOLTEN SALTS AND THE HOT WEAR (EROSION, ADHESIVE WEAR).